the problem is that once you deprotonate the carboxylic acid, intramolecular ring closing will be far more facile than hydroxide doing an SN2 at the primary bromide, so the linear substrate is just going to snap shut to the lactone. At best, you can expect a mixture of the lactone along with GHB, but I wouldn't count on it being a high yielding reaction. The 2 step protocol seems like a much safer choice to me.